# Beyond Predictive Fairness: Quantifying Attribution Consistency Across Demographic Groups in Diabetic Retinopathy Screening

arXiv: https://arxiv.org/abs/2608.18759 · HTML: https://mlsanity.com/paper/2608.18759/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 27 references that are in the map.
- First posted on arXiv 19 Aug 2026.

## Builds on (most cited first, 10 of 10)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Sanity Checks for Saliency Maps](https://mlsanity.com/paper/1810.03292/) (2018)
- [Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks](https://mlsanity.com/paper/1910.01279/) (2019)
- [Smooth Grad-CAM++: An Enhanced Inference Level Visualization Technique for Deep Convolutional Neural Network Models](https://mlsanity.com/paper/1908.01224/) (2019)
- [Fairness and Explainability: Bridging the Gap Towards Fair Model Explanations](https://mlsanity.com/paper/2212.03840/) (2022)
- [Evaluating the Fairness of Deep Learning Uncertainty Estimates in Medical Image Analysis](https://mlsanity.com/paper/2303.03242/) (2023)
- [Identifying the key components in ResNet-50 for diabetic retinopathy grading from fundus images: a systematic investigation](https://mlsanity.com/paper/2110.14160/) (2021)
- [This actually looks like that: Proto-BagNets for local and global interpretability-by-design](https://mlsanity.com/paper/2406.15168/) (2024)
- [Understanding challenges to the interpretation of disaggregated evaluations of algorithmic fairness](https://mlsanity.com/paper/2506.04193/) (2025)
- [On Biases in a UK Biobank-based Retinal Image Classification Model](https://mlsanity.com/paper/2408.02676/) (2024)

## Abstract (arXiv)

Fairness in medical imaging is commonly evaluated through subgroup performance metrics, yet it remains unclear whether models rely on consistent visual evidence across demographic groups. This work introduces the Explanation Consistency Score (ECS), a fairness-aware metric based on Jensen-Shannon divergence that quantifies the similarity of attribution maps across subgroups. Using diabetic retinopathy screening as a case study, ECS is evaluated globally and within disease severity. Experiments reveal that while predictive performance differs across ethnic groups, explanation consistency remains relatively high and shows no significant association with performance disparities. These findings suggest that predictive fairness and explanation consistency capture complementary dimensions of model behavior, motivating fairness evaluations that extend beyond predictive performance.